Optimizing the thickness profile on the calender: using roll bending and roll crossing correctly

Where the deflection comes from

In the roll nip of a calender, the polymer melt builds up pressures that, depending on recipe, temperature and nip geometry, are in the range of several hundred newtons per millimeter of roll width. This nip force deflects the rolls – not by much, but at film thicknesses of 100 to 300 µm, a few micrometers of deflection are enough to visibly distort the cross profile. Without countermeasures, the typical picture emerges: the film is thicker in the middle than at the edges.

Three levers are available: the roll crown (ground in permanently, suits only one operating point), roll bending (counter-bending via forces applied at the roll journals, infinitely adjustable) and roll crossing (the axes of the two nip rolls are rotated slightly against each other, which opens the nip toward the edges).

Why roll bending alone is not enough

Roll bending produces a parabolic correction; roll crossing produces a correction that increases more strongly toward the edges. The actual deflection follows neither curve exactly – it depends on the force distribution in the nip, i.e. on the rolling bank, web width and rheology. Only the combination of both levers makes it possible to bring the profile within tolerance across the entire width. In practice, often only one of the two levers is used; the result is residual errors at the edges or in the quarters of the web, which are then “solved” by edge trimming.

What happens during changeovers

When the target thickness changes, the nip gap, throughput and thus the nip force change as well. A changeover from 240 to 300 µm means not only a larger gap but a different force and temperature distribution – and therefore a different deflection. The roll bending and roll crossing values of the old operating point no longer fit. Anyone who “dials them in” on the running calender produces start-up scrap and loses time.

Calculating in advance instead of dialing in

Using the mass balance, momentum equation, heat balance and the rheological equation of state of the recipe, the flow lines, pressure curve and throughput in the nip can be calculated. From these follow the mean film thickness, the calendered width per roll, the expected shrinkage and – decisive for the profile – the nip force and its distribution. On this basis, the model delivers the roll bending and roll crossing values for minimum thickness tolerance at the new operating point.

The model can be used on a laptop or in the VR headset with images of your own line. This way, calender operators practice changeovers without occupying the producing line, and process engineers verify recipe or format changes before they are run. Case study: Changeover without start-up scrap.

Six checkpoints for the cross profile

1. Symmetry: Is the profile mirror-symmetrical left/right? If not, the cause is usually not roll bending or roll crossing but uneven temperature, rolling bank or web guidance.
2. Center vs. edge: Thick center, thin edges → deflection undercompensated. Thin center, thick edges → overcompensated.
3. Quarter errors: Waves in the quarters of the web indicate a mismatch between roll bending and roll crossing.
4. Temperature profile: Measure roll temperature across the width; a 2 °C difference changes the viscosity noticeably.
5. Rolling bank: Uniform bank size across the width? A one-sided bank produces one-sided nip forces.
6. Measuring system: Check calibration and sensor contamination before adjusting the line.

The complete list – including variations in the machine direction – is available as a downloadable checklist.